Overcoming the Political Exclusion of Migrants: Theory and Experimental Evidence from India - Cambridge ...
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
American Political Science Review (2021) 115, 4, 1129–1146
doi:10.1017/S0003055421000435 © The Author(s), 2021. Published by Cambridge University Press on behalf of the American Political
Science Association. This is an Open Access article, distributed under the terms of the Creative Commons Attribution licence (http://
creativecommons.org/licenses/by/4.0/), which permits unrestricted re-use, distribution, and reproduction in any medium, provided the
original work is properly cited.
Downloaded from https://www.cambridge.org/core. IP address: 46.4.80.155, on 24 Nov 2021 at 14:40:42, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0003055421000435
Overcoming the Political Exclusion of Migrants: Theory and
Experimental Evidence from India
NIKHAR GAIKWAD Columbia University
GARETH NELLIS University of California, San Diego
M
igrants are politically marginalized in cities of the developing world, participating in destination-
area elections less than do local-born residents. We theorize three reasons for this shortfall:
migrants’ socioeconomic links to origin regions, bureaucratic obstacles to enrollment that
disproportionately burden newcomers, and ostracism by antimigrant politicians. We randomized a
door-to-door drive to facilitate voter registration among internal migrants to two Indian cities. Ties to
origin regions do not predict willingness to become registered locally. Meanwhile, assistance in navigating
the electoral bureaucracy increased migrant registration rates by 24 percentage points and substantially
boosted next-election turnout. An additional treatment arm informed politicians about the drive in a subset
of localities; rather than ignoring new migrant voters, elites amplified campaign efforts in response. We
conclude that onerous registration requirements impede the political incorporation, and thus the well-
being, of migrant communities in fast-urbanizing settings. The findings also matter for assimilating
naturalized yet politically excluded cross-border immigrants.
C
ountries witnessing rapid economic develop- as social groups not exercising suffrage experience state
ment frequently struggle to assimilate new neglect (Fujiwara 2015).
migrants into cities. As the population of Brit- What accounts for migrants’ underrepresentation in
ain’s towns doubled in size during the industrial revo- politics? We theorize three mechanisms by which
lution, Friedrich Engels ([1845] 2010) described a mobility induces political marginalization. The first
burgeoning urban proletariat “cast out and ignored centers on migrants’ enduring economic and social ties
by the class in power” (114) and living in a “state of to their origin regions. Migrants who maintain close
dilapidation, discomfort, and misery” (viii). During the links to “home” may be unwilling to refocus their polit-
Great Migration in the United States, African Ameri- ical activities, opting to remain detached from political
cans escaping Jim Crow laws met with “unwritten, life in destination areas. Second, bureaucratic obstacles
mercurial, [and] opaque” resentment in northern cities; associated with participation in host regions—above
they were pushed to the margins, leading Martin all, the hassle costs of updating voter registration and
Luther King Jr. to lament that “Chicago has not turned navigating electoral bureaucracies—militate against
out to be the New Jerusalem” (Wilkerson 2010, 386). engaging there. Whereas governments normally assume
Internal migrants, who number one seventh of the responsibility for initiating the registration process in
world’s population, face similar challenges across much advanced industrialized states, we find that 16 of the
of today’s Global South (Bell and Charles-Edwards 20 most populous low- and middle-income democracies
2013). Political exclusion is commonplace (Bhavnani place the onus on citizens to initiate enrollment (see
and Lacina 2015; Thachil 2020; Weiner 1978). Evidence Supplementary Information A). Last, ostracism by local-
suggests that those who shift from the countryside to born residents and their elite representatives could
cities participate in destination-area politics at lower impede migrant integration (Dancygier 2010). “Sons of
rates than local-born residents.1 This shortfall matters the soil” parties that vilify newcomers have sprung up in
normatively, cutting against democracy’s promise of Mumbai, Karachi, and in parts of South Africa and
equal representation. It is also of practical consequence, Malaysia in recent decades (Bhavnani and Lacina
2018). It stands to reason that migrants meeting with
broad-based indifference or hostility will foresee few
benefits to sinking their energies on politics in host
Nikhar Gaikwad , Assistant Professor, Department of Political communities.
Science, Columbia University, nikhar.gaikwad@columbia.edu. We study the role played by these factors in under-
Gareth Nellis , Assistant Professor, Department of Political Sci- mining migrants’ political incorporation, which we con-
ence, University of California, San Diego, gnellis@ucsd.edu. ceptualize to comprise both citizen-side political
Received: May 13, 2020; revised: March 15, 2021; accepted: May 03, engagement and elites’ readiness to include citizens in
2021. First published online: June 08, 2021. their electoral coalitions. To do so, we fielded a large
randomized controlled trial in India, a leading case for
1
In this paper, “migrant” refers to internal migrants who share
citizenship and voting rights with local-born residents in destination
evaluating the political underrepresentation of internal
areas. “Destination area” refers to the jurisdiction to which migrants migrants. Our focus is on rural-to-urban migrants and
move, while “origin area” refers to the village, town, or city from the reasons why such individuals struggle to incorpor-
which they have relocated. ate politically in countries whose demographics are
1129Nikhar Gaikwad and Gareth Nellis
being transformed by high economic growth. We evalu- intimately bound to individual voting behavior. That
ate a door-to-door campaign to facilitate voter regis- migrants do not always assert their political participa-
tration among migrants to two cities: Delhi, the tion in their primary places of residence, despite pos-
national capital, and Lucknow, which is the capital of sessing the full constitutional rights to do so, poses a
Downloaded from https://www.cambridge.org/core. IP address: 46.4.80.155, on 24 Nov 2021 at 14:40:42, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0003055421000435
India’s largest state and is emblematic of a class of mid- significant puzzle.
tier cities increasingly attractive to jobseekers (Thachil Recent studies suggest that registration drives can in
2017). Partnering with an NGO in advance of the 2019 some cases be effective tools for spurring enrollment
national parliamentary elections, we recruited 2,306 among unregistered citizens (e.g., Harris, Kamindo, and
migrants who lacked local voter registration docu- van der Windt Forthcoming; Nickerson 2015). Yet to
ments. Half of those who expressed an interest in date there has been negligible theoretical or empirical
registering to vote in the city were then offered inten- work on the roadblocks to political access encountered
sive assistance in applying for a voter identification card by migrants in the developing world or, indeed, by
that enabled them to cast a ballot locally in the upcom- movers writ large. Going beyond prior studies, our
ing polls. In addition, we built a cluster-level experi- novel cluster-level experiment evaluates the “supply
ment on top of the individual-level design. Its purpose side” of political incorporation—investigating whether
was to inform politicians in a randomly chosen subset of politicians are responsive to news of the enfranchise-
neighborhoods that the registration drive had taken ment of a previously disempowered population group.
place, and thus to test whether attaining registered Observational studies on enfranchisement’s effects
status paved the way to migrants’ full-fledged political have generated mixed findings (cf. Paglayan 2021).
incorporation. Our layered research design enables us to experimen-
Previewing the results, there is little to suggest that tally probe equilibrium dynamics as new groups of
migrants’ ongoing links to their former places of resi- voters enter the electoral fray and politicians update
dence prevent them from incorporating politically at their campaign strategies in reaction. To the extent that
their destinations—our first theoretical conjecture. these strategies entail the targeting of individualized
Asked whether they wished to register locally, 98% of benefits and local public goods, our study can further
eligible respondents replied “yes,” indicating that vol- advance understanding of the sources of urban depriv-
untary disengagement is rare in our sample. This is ation (Auerbach 2019).
striking because interviewed migrants reported signifi- Finally, as we document, naturalized cross-border
cant social and economic ties to their prior hometowns. immigrants consistently vote at lower rates than
By contrast, there is clear evidence to support our native-born citizens in wealthier democracies. Across
second theoretical claim: that bureaucratic obstacles countries in the Organisation for Economic
to registering to vote hinder migrants’ electoral partici- Co-operation and Development, turnout rates aver-
pation. Alleviating these constraints—by providing aged 80% for native-born citizens compared with
at-home assistance in completing and submitting voter 74% for foreign-born naturalized citizens between
registration documents—increased migrant registra- 2008 and 2016 (OECD 2019, 128). An analogous
tion rates by 24 percentage points and next-election 10 to 12 percentage point disparity existed in the United
turnout by 20 percentage points. It also shifted down- States throughout the 2000s (Wang 2013). Analyzing
stream outcomes, raising political interest and percep- data from the 2014 World Values Survey, which
tions of local political accountability. covered 52 countries, we find that 82% of native-born
Does elite nonresponsiveness further undermine citizens report that they “always voted” or “usually
migrants’ local political incorporation, as our final the- voted,” compared with 71% of foreign-born citizens
oretical proposition predicts? The eagerness of (see Online Appendix A). Civil society organizations
migrants to accept registration assistance suggests that advocating on behalf of Hispanic communities in the
anticipation of ostracism is not a major determinant of United States and Muslim communities in Europe, for
exclusion on the demand side. Yet, we go further to example, have underscored the importance of making
assess experimentally whether the basis for such per- these groups’ voices heard in the political arena.2 Nat-
ceptions exists. If city politicians are constrained by the uralized immigrant groups may therefore also gain
antimigrant preferences of urban electorates, then from interventions that are similar to the one evaluated
learning about the mass registration of migrant voters here (Braconnier, Dormagen, and Pons 2017; Pons and
locally should fail to influence their campaign strat- Liegey 2019).
egies. Against this expectation, we find that election-
eering increased in the vicinity of polling stations listed
in our communications to candidates. As migrants
found a place on local electoral rolls and politicians CONCEPTUALIZING MIGRANT POLITICAL
learned as much, candidates began soliciting migrant INCORPORATION
support. Overall, we conclude that stringent registra-
tion requirements—rather than “opting out” or expect- Before presenting our theory, we introduce the concept
ations of ostracism by local political machines—drive of political incorporation and highlight its relevance for
the political incorporation gap between migrants and migrants worldwide.
local-born residents in the fast-growing cities we study.
Our study breaks new ground. We study a “patron-
2
age democracy” where access to government benefits is See, for example, the work of Voto Latino (bit.ly/3bBZjsC).
1130Overcoming the Political Exclusion of Migrants: Theory and Experimental Evidence from India
Critical to our conceptualization is the dual contri- marshal case study evidence from Côte d’Ivoire and
bution of demand- and supply-side inputs in bringing Indonesia, where, according to observers, politicians
about migrant incorporation.3 On the demand side, have weaponized sons-of-the-soil narratives to “turn
incorporation requires migrants’ full and active partici- the politics of resentment to their electoral advantage.”
Downloaded from https://www.cambridge.org/core. IP address: 46.4.80.155, on 24 Nov 2021 at 14:40:42, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0003055421000435
pation in destination-area politics. Behaviorally, it pre- Weiner (1978, 9) describes elites’ antimigrant stances in
sumes that migrants first register to vote in city-based the cities of Eastern Europe following the collapse of
elections—a prerequisite for subsequent participation the Hapsburg Empire.
in formal political institutions—and then turn out to In short, the historical and comparative case litera-
cast a ballot, the fundamental democratic act. Attitud- tures bear out the two-sided barriers to political incorp-
inally, it entails shifts pertaining to political interest, oration for migrants. Our next task is to explain what
perceptions of political accountability, efficacy, and drives variation in this phenomenon.
trust.4 Yet citizen action alone is insufficient to bring
about full political incorporation, from our perspective.
For that to happen, political elites must reciprocate by
treating migrants equally vis-à-vis local-born residents THEORIZING THE MIGRANT–LOCAL
and by acknowledging them to be bona fide members PARTICIPATION GAP
of local electorates. We minimally expect incorporation We now theorize the key citizen- and politician-side
to include outreach to migrant citizens during campaign constraints to migrants’ political incorporation.
seasons, inclusion of migrants in local networks of Regarding citizens’ constraints, we develop two main
clientelistic exchange, and sensitivity to problems faced hypotheses—one centered on voluntaristic detach-
by migrant communities. Paying attention to these ment, the other on bureaucratic hurdles—explaining
supply-side actions is essential, not the least because nonincorporation. These hypotheses hew to a cost-
politicians’ failure to include migrants in local political benefit analysis of political engagement, which posits
coalitions may demotivate migrants from taking steps that citizens participate politically when the expected
to participate in the first place. benefits of doing so exceed the expected costs (Downs
Our conceptualization maps onto the empirical rec- 1957).5 Overlooked in standard models, however, are
ord. In the realm of domestic migration, there is sig- costs and benefits that fall uniquely on those who move.
nificant evidence that mobility is associated with lower Regarding politicians’ constraints, our third hypothesis
political incorporation globally. Changing place of resi- homes in on the electoral pressures felt by candidates to
dence in Costa Rica “disrupts” voting and is associated eschew migrants. Such neglectful treatment by elites
with an eight to nine percentage point reduction in has the potential to feed back into migrants’ decisions
turnout propensity (Alfaro-Redondo 2016, 73). Focus- to engage politically in the city.
ing on Turkish municipalities, Akarca and Tansel
(2015) estimate a strong negative province-level rela-
tionship between in-migration and electoral participa- Voluntary Detachment
tion. Gay (2012) shows for the United States that use of
a randomly assigned housing-relocation voucher Migrants may voluntarily decline to engage in
reduced the probability of voting in national elections destination-region politics because they see greater
by seven percentage points. Qualitatively, scholars advantages to remaining politically involved in their
have substantiated internal migrants’ relative political home regions. Unlike local-born residents, migrants
disengagement in Nigeria, Colombia, Ukraine, and possess a choice about where to exercise their political
Myanmar (see Supplementary Information B). Ample participation: they can do so either in their region of
evidence also attests to politicians’ exclusionary origin—their default option—or in the place they settle.
behaviors—that is, the core supply-side impediment Thus, one engagement repertoire for migrants entails
to incorporation. Côté and Mitchell (2016, 662–66) delinking their place of residence (for clarity of expos-
ition, “the city”) from their place of political participa-
tion (“the village”).
A delinked engagement strategy holds out several
3
We parallel the “multidimensional view of integration” advanced
by Sobolewska, Galandini, and Lessard-Phillips (2017), who note the
attractions for migrants. First, social and emotional
panoply of cultural, social, and economic factors influencing the attachments weigh against withdrawing from politics
environment in which migrants engage politically. in origin areas. Political interest develops during the
4
The attitudinal components of incorporation may be spelled out early, formative years of individuals’ lives and does so
further. First, healthy democracy rests on a watchful and informed
electorate; thus, we conceptualize politically incorporated migrants
5
to be those who take an interest in politics at the local and national Beyond this rationalist paradigm, others have pinpointed the influ-
levels. Second, enfranchised individuals may come to perceive polit- ences of habit formation (Meredith 2009), life-cycle timing
ical elites as more subject to citizen control and, consequently, as (Ferwerda, Finseraas, and Bergh 2020), and economic self-
disincentivized to engage in corruption or mismanagement. For this investments (Peters, Schmeets, and Vink 2019) on individual political
reason, migrants’ assessment of local political accountability matters engagement. Voicu and Comşa (2014) view immigrants’ voting pro-
for incorporation. Third, incorporation implies enhanced political pensity as a product of socialization in both destination and origin
efficacy: the sense that “people like me” have influence over the areas; by seeing political participation as a socially embedded act,
government. And fourth, a high degree of trust in political institu- their culturalist approach elucidates group-level differences in
tions, reflecting openness to giving up some personal autonomy to the migrant engagement whereas our focus is on individual-level con-
state, is intrinsic to the notion of political incorporation. straints.
1131Nikhar Gaikwad and Gareth Nellis
in specific locales—places where social pressures to Windt Forthcoming; Nickerson 2015).8 Missing from
participate may be felt more intensely. At the same consideration in this literature have been the particular
time, migrants often find themselves socially isolated challenges migrants confront as they grapple with elect-
at their destinations. Among rural-to-urban migrants in oral bureaucracies.
Downloaded from https://www.cambridge.org/core. IP address: 46.4.80.155, on 24 Nov 2021 at 14:40:42, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0003055421000435
China, for example, “non-kin social ties between Why might migrants be asymmetrically hard hit by
migrants and local urban residents are limited [and] bureaucratic registration hurdles? Migrants’ difficulties
non-resident ties still make up the majority of migrant may be the unintended consequence of systems
networks” (Yue et al. 2013, 1720). This is true of cross- designed with nonmovers in mind. First, newcomers
border immigrants too.6 Such dislocation effects are lack familiarity with local government procedures,
likely to be increasing in the cultural distance between rules, and regulations. The everyday knowledge
migrants and nonmigrants. Thus, social relationships— required to navigate local bureaucracies (e.g., knowing
that are strong in origin areas and weaker in destinations the location of the nearest ward office) is likely to be
—may prolong origin-area political participation, even second nature to local-born residents but opaque to
after migrants move away. outsiders. Second, in multilingual contexts, migrants
Second, economic motivations could dissuade from peripheral regions often speak a different lan-
migrants from shifting their locus of political participa- guage or dialect from those living in urban centers.
tion. Rationally, migrants with significant material assets Migrants who struggle with foreign-language forms will
(e.g., property) in their former homes will seek to main- be handicapped in their bid to register to vote. Third,
tain a political voice there postmigration—to ward off registrants must typically provide supporting docu-
expropriation and other economic threats. Intangible ments along with their application. Here, too, migrants
assets provoke similar calculations. Notably, migrants lag—particularly those living in informal settlements
may be reluctant to sever long-nurtured clientelistic without title deeds or formal utilities. Finally, in many
relationships with origin-area politicians. These relation- contexts, migrants are shouldered with a “double-
ships profit election-minded politicians, too, who have registration burden”: they are required to deregister
accordingly been known to bus migrant workers back to to vote in their prior place of residence before reregis-
villages at election time and to offer gifts and additional tering in their destination region. Locals do not have to
benefits to woo migrants back during polls.7 Analo- jump through this additional hoop.
gously in the international domain, Wellman (2021) Alternatively, bureaucratic elites may deliberately
finds that ruling parties encourage diaspora voting when erect barriers to thwart migrants seeking to register.
they perceive electoral advantages to doing so. Bureaucratic bias against culturally distinct outsiders—
Summing up, there are compelling reasons for or against marginalized groups overrepresented in the
migrants to wish to anchor their political participation migrant pool—can amplify enrollment costs (White,
in origin regions, rather than to transfer it to their point Nathan, and Faller 2015). Bureaucrats may recognize
of destination, offering an explanation for the migrant– migrants and treat them in a demeaning or dismissive
local incorporation gap. This leads us to hypothesize way during in-person interactions at government
that migrants who are more socially and economically offices. When processing documents submitted
attached to their places of origin will be more likely to remotely, bureaucrats can use proxies such as ethnic
remain politically detached in their new places of resi- naming conventions or addresses (along with informa-
dence. tion about neighborhood demography) to detect
migrant status. Bureaucrats can then drag their feet,
refuse advice, call for supplemental evidentiary docu-
Bureaucratic Hassle Costs ments, or deny migrant petitions on spurious grounds.
A second explanation for migrants’ political disengage- Engaging with the state is demanding for almost any
ment emphasizes the high administrative barriers class of citizens. The foregoing discussion highlights the
migrants face in registering to vote in destination regions. theoretical reasons why migrants may be levied with a
Registration involves gathering and copying paperwork registration surcharge, due both to the side effects of
and completing and submitting forms. Citizens routinely procedures created without thought to “voters on the
procrastinate on time-consuming bureaucratic tasks. move” and willful attempts to suppress migrant incorp-
Randomized trials of voter registration campaigns in oration by bureaucrats. A testable implication is that
France, Kenya, and the United States demonstrate that programming crafted to ease the voter registration
enrollment is sometimes—though not always—sensitive costs borne by movers will have pronounced, positive
to convenience costs for the average citizen (Braconnier, effects on their subsequent political incorporation.
Dormagen, and Pons 2017; Harris, Kamindo, and van der
Political Ostracism
6
Because “migrants cluster in ethnic communities” and have
“limited contact with the host society,” their “connection to the Our final theoretical perspective attributes migrants’
political life in their host country is at best limited” (Careja and political exclusion to ostracism and antimigrant
Emmenegger 2012, 880–81), whereas their attachment to country-of-
origin politics remains strong (Alarian and Goodman 2017, 140). See
8
also Fouka (2019). To contextualize our study, Supplementary Information C provides
7
“MP Polls: Migration of Voters Big Worry for Parties in Bundelk- a systematic review of field-experimental studies of the effects of
hand Region.” Economic Times, November 24, 2018. voter-registration assistance published to date.
1132Overcoming the Political Exclusion of Migrants: Theory and Experimental Evidence from India
backlash by urban elites. Across domains, local-born transparent to elites, therefore, it makes sense to focus
residents foresee that the arrival of newcomers will recruitment and mobilization efforts on locals.
heighten labor market competition, strain public ser- If it is the case that ostracism by politicians is what
vices, and “dilute” the ethnocultural fabric of urban drives migrant nonincorporation, a testable implication
Downloaded from https://www.cambridge.org/core. IP address: 46.4.80.155, on 24 Nov 2021 at 14:40:42, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0003055421000435
society (Gaikwad and Nellis 2017; Scheve and Slaugh- follows: even upon learning that migrants are regis-
ter 2001). The upshot is that locals, along with their tered to vote locally, politicians will decline to bring
elected representatives, are prone to exhibit antimi- migrants into their local electoral coalitions. This is the
grant attitudes. This can manifest as passive indiffer- third hypothesis we set up our empirical approach to
ence: migrant entry is permitted but nothing more is evaluate.
done to smooth migrants’ integration or to deliver them
services to which they are entitled. It can also materi-
alize as active antagonism—for instance, through the
passage of voter suppression laws.9 These strategies are MIGRATION AND VOTING IN INDIA
especially likely to arise under conditions of local eco-
We now describe a study setting—India—conducive to a
nomic scarcity (Dancygier 2010). In short, politicians
rigorous test of our three theoretical claims. India has 1.4
can respond with apathy to enfranchised migrants or, in
billion citizens, 900 million eligible voters, and an esti-
certain circumstances, by stifling migrant turnout.
mated 325 million internal migrants, comprising 29% of
Migrants may disengage politically in these environ-
the country’s population (Government of India 2010).
ments. Those who expect to be sidelined will anticipate
At 34%, India’s current level of urbanization is low by
little gain from taking part in city politics. Migrants
international standards. However, the country’s urban
fearful of police harassment, additional taxes, and pog-
population is projected to grow to 590 million people by
roms may prefer to live in the “shadow of the state.”
2030, up from 290 million in 2001, and the preponder-
Ethnographic accounts attest to these concerns. For
ance of all new jobs generated over the next decade will
example, Jones (2020, 93) finds that migrants in Guar-
be city-based (Sankhe et al. 2010).
iba, Brazil “do not belong and are not entitled to make
India’s rural-to-urban migrants constitute a disadvan-
demands” on the government because “civic ostracism
taged population category. Online Appendix B analyzes
places them on the margins of local politics [and] they
nationally representative survey data, revealing that
have internalized the exclusion that they experience at
migrant-engaged households are poorer than nonmi-
the hands of permanent residents.” How general these
grant households overall and they are more likely to
experiences are remains to be established.
belong to marginalized ethnic communities—differences
There is also a subtler logic by which politicians’
that underline the integration obstacles that internal
uncertain beliefs about migrants’ preferences can prod-
migrants face.10 A United Nations report notes that “a
uce exclusion. In places where antimigrant sentiment is
holistic approach is yet to be put in place that can address
muted, it seems intuitive that parties would help
the challenges associated with internal migration in
migrants register and participate. Where parties can
India” (UNESCO 2012, 2). New migrants encounter
be confident of migrants’ probable vote choice, such
discrimination in accessing government services
party-led drives to enlist migrant voters make rational
(Gaikwad and Nellis 2021a), migrant slums lack basic
sense. Yet, more commonly, migrants are unknown
facilities (Auerbach 2019), and demands for “locals only”
quantities in local politicians’ eyes. Ethnocultural dis-
employment quotas and discriminatory language stipu-
similarities between politicians and migrants make
lations have been prevalent (Weiner 1978).
their partisan leanings hard to discern; the tendency
The migrant/local participation gap is stark in India.
for low-income migrants to reside in dense, heteroge-
Subnational states with more migrants evince lower
neous informal settlements renders their political pref-
voter turnout (Tata Institute of Social Sciences 2015,
erences less legible to urban elites; and politicians may
30). Survey data from five states suggest that between
presume ex ante that migrants’ home attachments will
60% and 83% of domestic migrants did not cast a ballot
depress demand for city-based registration. Mean-
in at least one national, state, or local election after
while, the costs of shepherding migrants through the
moving (Tata Institute of Social Sciences 2015, 32).
process are high. Running migrant-focused registration
Representative surveys collected after the 2014
drives thus carries two risks for politicians and parties:
national election revealed turnout to be 69% in rural
(a) newly registered migrants may not turn out to vote,
constituencies, 63% in smaller cities and towns, and
meaning that scarce resources have been squandered,
57% in large metropolitan areas (Kumar and Banerjee
and (b) migrants may accept registration help but then
2017, 83).11 Micro-level studies in Delhi identify low
go on to vote for a competitor. To the extent that the
political preferences of local-born residents are more
10
According to official data, 60% of migrants moving from neigh-
boring states and 84% of migrants moving from non-neighboring
states do not share a common language with local-born residents in
destination areas; among urban-destination migrants, 61% of women
9
A historical example of migrant-targeted voter suppression is pro- cite marriage and 56% of men cite employment as the primary reason
vided by the French Second Republic, which in May 1850 instituted a for migration (Government of India 2010; Kone et al. 2018).
11
local residency requirement that disenfranchised city-based workers In Online Appendix B.4, we present evidence that the urban–rural
from the countryside and “drove republican politics underground” participation gap persists after controlling for socioeconomic status.
(Berenson 1984, 169). That rural turnout exceeds urban turnout in India deepens the puzzle
1133Nikhar Gaikwad and Gareth Nellis
turnout among urban-based migrants as a prime con- on prior migration work by exploiting the paired com-
tributor to this rural–urban turnout divide: in 2014, parison of Delhi and Lucknow (see Thachil 2017).
only 65% of recent migrants to Delhi possessed a voter
ID card allowing them to vote in city elections, whereas
Downloaded from https://www.cambridge.org/core. IP address: 46.4.80.155, on 24 Nov 2021 at 14:40:42, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0003055421000435
the overall average for Delhi residents was 85%.12 Local Context
Similarly, Thachil (2017) finds in a sample of Delhi
construction workers that only one in five migrants had To contextualize our study, we document the formal
ever voted in the city’s elections. A survey of India’s voting registration process in India and provide prelim-
five largest cities conducted after the 2019 general inary insights—from prior research and qualitative
elections revealed that 91% of urban-based migrants fieldwork—illuminating the difficulties migrants
in their twenties reported that they were not registered encounter in this regard.
to vote in the city.13 The de facto disenfranchisement of
internal migrants has been dubbed a “serious infirmity
in the electoral process of the world’s largest Voter Registration Process
democracy.”14 India uses simple plurality rules to elect 543 Members
of Parliament (MPs) to single-member districts. Each
Case Selection citizen in our sample is further represented by an
elected Member of the Legislative Assembly (MLA,
Our study was fielded in two cities. Delhi is home to a state-level position) as well as an elected municipal
19 million residents and is India’s national capital; it corporator.
absorbs more migrants than any other metropolitan Our investigation was timed to coincide with the 2019
area. Like other megacities, its urban landscape is Indian national (MP) elections. Citizens are required to
dotted with jhuggi jhopri (“slum hut”) dwellings, con- initiate the registration process, which can be done
structed from plastic, corrugated iron, and wood. Esti- online or on paper. The process, which is mandated
mates suggest that 40% of Delhi’s population by the federal Election Commission of India and is the
comprises migrants from other Indian states.15 Luck- same nationally, has four components:
now, meanwhile, has 2.8 million residents. It is the
capital of Uttar Pradesh, India’s largest state. Most • Form-filling. Registrants must complete the “Appli-
migrants to Lucknow are attracted by higher wages cation for Inclusion of Name in Electoral Roll”
and originate from economically backward districts of (Form 6, reproduced in Online Appendix C).
the state (Bose and Rai 2014, 53–55). Migrants must also annul their registration in their
Delhi and Lucknow were chosen for three reasons. prior place of residence and submit the deletion slip
First, we heeded the call from scholars of urban politics (Form 7), or they must declare that they were not
to study not only megacities but also small and medium- previously registered elsewhere.
sized urbanities, which, as Auerbach et al. (2018, 262) • Documentation. Citizens must provide proof of local
emphasize, are the world’s “most quickly expanding residence (for example, a locally addressed electric/
urban centers.” The comparison of Delhi, a Tier I gas bill, ration card, driver’s license, or bank pass-
Indian city, and Lucknow, a Tier II city, adds crucial book), passport size photographs, and proof of age
generalizability in this respect. Second, and at the same (e.g., birth certificate). If the applicant is a tenant, her
time, we sought a controlled comparison, opting for two landlord is usually expected to sign an affidavit con-
locations that were broadly similar in terms of the firming current occupancy.
locally dominant language used (Hindi), historical • Submission. Applications are submitted to the regis-
and colonial legacies, national electoral preferences, tration office of the Assembly Constituency in which
socioreligious structures, and the primary sending the citizen resides. Alternatively, applicants can hand
regions of their migrant populations.16 Last, we built over the documents to a Booth Level Officer (BLO)
during brief voter registration drives conducted
annually by local election offices.
of low rates of migrant participation in destination cities insofar as • Verification. Local voter registration offices process
socialization in high-turnout origins areas is not evident in migrants’
city-based political behaviors.
the forms. If in order, a BLO will then pay an
12
See Kumar and Banerjee (2017). In a 2018 survey of 6,884 slum in-person visit to the applicant’s given address to
dwellers in Mumbai, 61% of respondents who were born out of state verify that the submitted photograph matches the
reported having a local voter ID card, whereas the corresponding applicant. A voter identification card is mailed by
figure for in-state respondents was 71%. Just 25% of migrants post to the applicant following approval, and their
reported having an origin-place voter ID card (Gaikwad, Nellis, name is entered on the local voter roll. Rejections
and Thomas 2021).
occur either when documents are judged incomplete
13
“91% of Urban Migrants Not Registered as Voters in Cities They
Live.” Hindu BusinessLine, March 6, 2019. or improperly filled or when the applicant is not
14
“The Migrants Indian Democracy Forgot.” Diplomat, February found during the BLO’s visit.
7, 2019.
15
“Delhi Has Highest Share of Inter-state Migrants.” Hindustan
Times, July 28, 2019.
16
Note that although the main sending regions for these cities are dialects make newcomers’ migration status easily linguistically visible
also predominantly Hindi-speaking, highly localized differences in to local-born residents.
1134Overcoming the Political Exclusion of Migrants: Theory and Experimental Evidence from India
Barriers to Registration Ambedkar Camp … clueless and ill-prepared.”21
According to another, “A few years ago, I had
Inertia, corruption, and classism mar India’s voter regis-
approached the Election Officer at the Dwarka voter
tration system. Peisakhin (2012) found the median pro-
registration office only to be told that the concerned BLO
Downloaded from https://www.cambridge.org/core. IP address: 46.4.80.155, on 24 Nov 2021 at 14:40:42, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0003055421000435
cessing time for new voter registrations to be 331 days
had resigned and there was nobody assigned to our basti
for slum residents in Delhi. To maximize rents, “officials
at the moment.”22 These accounts testify to migrants’
at election registration offices did almost everything in
predicaments in registering to vote in destination cities.
their power to indirectly encourage applicants … to turn
Indian parties have, at times, engaged in voter regis-
to middlemen for assistance” (Peisakhin 2012, 139).
tration drives to bring on board new supporters. Yet
Nonbribe payers were additionally harassed, being
parties’ record of migrant outreach is patchier. Regis-
asked to supply documents not required by law.
tration costs are high for parties because Indian elect-
Our own interviews back these claims, while also
oral law requires that registration applications be filed
highlighting the peculiar problems migrants face in
before candidate nominations are declared, at a time
registering to vote.17 The BLOs betrayed animus
when party organizations are typically shuttered.
toward migrants in interviews, referring to them as
Knowing whether and in what way migrants will vote
“troublemakers.” According to one officer:
is challenging: India is multiethnic and multilingual; it
has sprawling migrant-dominant slum settlements, and
Petty crimes in the area have shot up since the recent
voters and candidates routinely switch parties. Online
influx of migrants. They use voter IDs to get loans and
Appendix D analyzes representative data from the
then abscond. I would perform numerous background
checks on a prospective tenant who is a migrant since all
2015 Delhi state assembly elections showing that recent
his ID proofs will carry my address, and it is me who internal migrants were 13.6 percentage points less
stands to get bogged down by all the police paperwork in likely to have been targeted for voter outreach by the
the event of an untoward incident. At the voter office, we city’s political parties compared with Delhi’s long-term
are advised to exercise caution in our dealings with residents. Overall, India’s urban migrants evince low
migrants.18 rates of registration under the status quo. Parties’
engagement with migrants appears to fall short of their
engagement with local-born residents.
Local elites were skeptical about migrants’ motives for
registering. “It wouldn’t be a stretch to say that in the
case of migrants, the primary motive for obtaining
voter IDs is not the right to vote itself;” rather, they RESEARCH DESIGN
care only about the “potential benefits,” including a
We implemented a large, multilevel field experiment to
“claim to a government plot” if the slum is demolished,
shed light on the reasons why migrants often do not
as well as “healthcare and education benefits.”19 Prad-
incorporate politically into destination cities and the
hans (local community leaders) flagged landlords’
degree to which such disengagement is remediable.
apprehensions about migrants obtaining local voter
ID cards:
Sampling, Recruitment, and Baseline Survey
Landlords frequently refuse to sign the mandatory under-
taking required by a tenant while filling registration Form
Sampling and the administration of the baseline survey
6. Further, landlords have, in the past, dragged the election proceeded in four stages.
office to court for registering migrants without their
approval. Subsequently, BLOs have been as wary as the 1. Site selection. We first generated a list of 150 migrant-
landlords themselves in dealing with tenants in bastis [slum dominated settlements, which is to our knowledge the
colonies].20 most comprehensive list of such settlements for Delhi
and Lucknow. To make this initial frame of potential
sites, we relied on census data, schedules of informal
Migrants harshly criticized the election management
settlements produced by city governments, and
system. “The voter office is jolted out of its inactivity
information gleaned from respondents at migrant-
only days before the elections. I suspect bastis are not
dominated labor chawks (markets). Scoping teams
even a priority for them. This year BLOs arrived at
of 30 researchers assessed the suitability of potential
sites over a period of nine months. During visits to
settlements, they surveyed residents regarding the
17
Confidential, semistructured interviews with migrants and local possession of local voter ID cards. The study includes
elites were conducted outside the sample employed for the random- those settlements where informants reported the
ized trial described below. Individuals were approached based on
snowball sampling methods. The purpose of this field research was to
greatest numbers of unregistered internal migrants
provide a textured understanding of the local context and to cast light residing.
on causal mechanisms. 2. Individual-level screening. Within selected neigh-
18
Interview: BLO, Karol Bagh Assembly Constituency, Delhi, borhoods, enumerators employed interval sampling
August 22, 2019.
19
Interview: BLO, New Delhi Assembly Constituency, Delhi,
21
August 13, 2019. Interview: Resident, Karol Bagh, Delhi, September 12, 2019.
20 22
Interview: Pradhan, Punjabi Bagh, Delhi, September 7, 2019. Interview: Pradhan, Sagarpur, Delhi, September 22, 2019.
1135Nikhar Gaikwad and Gareth Nellis
to choose potential households to interview. hypothesis regarding bureaucratic hassle costs. Simple
Informed consent was requested from the adult randomization was used to assign individuals who com-
household respondent with the next upcoming birth- pleted the long baseline survey to T1 with 50% prob-
day. Therefore, only one subject was sampled per ability. Remaining subjects were assigned to a pure
Downloaded from https://www.cambridge.org/core. IP address: 46.4.80.155, on 24 Nov 2021 at 14:40:42, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0003055421000435
household. Subjects stating that they neither were control condition.
born in the city nor had a voter ID card enabling The intervention was an intensive door-to-door facili-
them to vote there were deemed eligible. Thus our tation campaign to help migrants obtain a local voter
full sample is intended to be representative of identification card (see Online Appendix C for further
unregistered migrants in sampled colonies. details). This card would enable them to participate in
3. Short baseline on omnibus sample. Eligible subjects the forthcoming national elections in the city where they
were asked about basic demographics; their past were living (either Delhi or Lucknow). To begin, a
voting behavior; and the extent of their social, eco- worker trained with the help of our NGO partner visited
nomic, and political connections to their home vil- migrants at their place of residence and presented their
lages or towns. At the conclusion of this module, we credentials. The worker described the benefits of hold-
asked whether they wished to obtain a local voter ID ing a local voter ID card, the process for getting one, and
card. The survey ended if a respondent replied “no.” the type of assistance the worker could provide. The
These initial questions, as well as the overall take-up worker asked whether the migrant had the necessary
rate, are used to test the plausibility of the voluntary supporting documents at hand.25 If they did, the migrant
detachment hypothesis. was asked to gather those documents in time for a future
4. Long baseline on experimental sample. Respondents visit and a time was set. Migrants who lacked these
replying “yes” entered the experimental sample. We documents were instructed on how to get them.
posed a larger set of demographic and attitudinal At the follow-up visit, the worker helped the migrant
questions to this group. Descriptive statistics are to complete the required forms online using an
provided in Supplementary Information F. These internet-enabled computer tablet. At the end of the
subjects were then randomized into different treat- meeting, workers ensured that the form, along with
ment conditions.23 uploaded photographs of the required documents,
were submitted to the local registration office. The
Representativeness worker then tracked the progress of the applications.
Where problems arose—often, for example, BLOs
In Online Appendix B, we benchmark our experimen- were unable to track down the applicant at their listed
tal sample against nationally representative survey data residence—the worker intervened to try to fix the issue.
collected in the second round of the Indian Human
Development Survey (IHDS-II). Comparing household
characteristics in IHDS-II by migrant/nonmigrant status Treatment 2: Cluster-Level Information
in both rural and urban areas, we highlight three key Dissemination Campaign (Politician-Targeted)
respects in which our sample generalizes. First, migrants
overall evidence lower political participation rates than Our second treatment arm—henceforth, “T2”—operated
nonmigrants. Second, marginalized caste groups and the at a cluster level and was intended to help adjudicate
poor are overrepresented both in our sample and in the the plausibility of our third theoretical claim concerning
wider internal migrant population. A third point of note politician ostracism. Using GIS software and publicly
concerns connectivity to patronage networks. Conceiv- available information on the locations of polling sta-
ably, our sampling criteria, which hinge on migrants’ tions in Delhi and Lucknow, we identified the 87 polling
stated willingness to register, select for migrants who are stations in closest proximity to our geo-located sample
comparatively excluded from clientelistic channels in of experimental subjects. Each subject was tagged to
home regions. In fact, the IHDS-II benchmark demon- the nearest one of these polling stations. We then
strates that rural migrant-sending households are less divided the 87 polling stations into four blocks, defined
enmeshed in client–patron relationships than were by city and whether the number of experimental sub-
others, potentially spurring migrants’ decision to “move jects assigned to that polling station was above or below
to opportunity” to begin with.24 the city’s median number of experimental subjects per
polling station. Last, we randomly assigned polling
stations within each randomization block to either the
Treatment 1: Individual-Level Registration T2 intervention or to the T2 control condition.
Drive (Migrant-Targeted)
Our primary intervention (“T1”) operated at the level
25
of individual migrants. Its purpose is to test our second Our analysis of the nationally representative IHDS-II data reveals
that migrant households were substantially less likely than were
nonmigrant households to possess officially acceptable proof-of-
23
Note, therefore, that the “experimental” sample is a subset of the residence and photo ID documentation (see Online Appendix B.5).
“omnibus” sample, comprising those individuals who wished to This finding was buttressed in our ethnographic interviews. Accord-
register to vote in their destination cities. The conclusions from our ing to one interview respondent, “At the time of applying, migrants
randomized study thus hold for this population of interest. need to attach an ID proof to the form. Some do not even possess the
24
According to the IHDS-II sample, approximately 9% of rural bare minimum proof and have to bribe the BLOs.” Interview:
households engaged in seasonal migration over the past five years. Pradhan, Indira Colony, Delhi, September 2, 2019.
1136Overcoming the Political Exclusion of Migrants: Theory and Experimental Evidence from India
Randomization to T2 was thus fully independent of the to the T2 treatment and 0 if assigned to T2 control, and
T1 trial. X is a vector of controls—the same set used for the T1
Between two and four weeks before election day— estimation—and δs are block fixed effects. We use
during India’s month-long campaign season and cluster-robust standard errors, clustering at the polling
Downloaded from https://www.cambridge.org/core. IP address: 46.4.80.155, on 24 Nov 2021 at 14:40:42, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0003055421000435
following parties’ selection of candidates—we sent booth level, which is the unit of randomization. As was
individually tailored letters, WhatsApp messages, and prespecified, we reweight individuals such that each
emails to four types of politicians tied to each polling cluster contributes equally to the estimation.
booth: the incumbent MP, officially declared MP can- The intent-to-treat (ITT) effect is the quantity of
didates, incumbent MLAs, and incumbent municipal interest across all experimental analyses. In tests of
corporators. Politicians’ contact details were gleaned balance and attrition we statistically find no threats to
from public databases. The messages were written in internal validity (see Online Appendix I). We success-
both English and Hindi. They informed the politicians fully reinterviewed 91% of baseline subjects at endline.
that a voter registration drive had been recently carried Our study was preregistered at the Evidence in Gov-
out among internal migrant communities in their area ernance and Politics registry (#20191007AA).28
(see Online Appendix E for examples of the commu-
nications).26 The purpose of the treatment was to dis-
seminate information about the registration drives that Ethics
had taken place and thus to positively update politi- Our experimental design was informed by ethical best
cians’ beliefs about the average registration rate of practices. To begin, we selected a sample size that was no
migrants in those localities. larger than we deemed necessary to detect meaningfully
sized treatment effects. Further, the sites in which we
Outcomes worked were not electorally competitive—the national
ruling party secured strong victories across all study
Outcomes were measured in an endline survey con- constituencies in the preceding election and in the elec-
ducted approximately two months after the elections tion we investigated—implying that there was a negli-
were held. The study timeline is shown in Online Appen- gible risk of affecting aggregate electoral outcomes.
dix F, and the survey question wordings are given in Next, our nonpartisan intervention partner had
Supplementary Information G. Methods used to create already conducted extensive voter registration cam-
indexed outcomes are described in Online Appendix G. paigns among urban migrants. It is a not-for-profit
organization working in the fields of slum rehabilita-
tion, housing rights for migrant workers, and slum
Analysis sanitation and public health. Public advocacy for
For estimations involving T1, we run ordinary least migrants has been at the core of its activities, meaning
squares regressions of the following form: that such drives—conducted both by our partner and
other civil society organizations—would have occurred
Yi ¼ a þ b T1i þ X0 i λ þ ui , (1) in our absence.29 This was a multistakeholder project.
As academic partners, our role was to scientifically
where, i indexes subjects, Y is the dependent variable of evaluate the efficacy of the NGO’s efforts in helping
interest, T1 is a dummy taking 1 if the subject was empower migrants to take up their constitutional rights
assigned to the registration facilitation campaign and to participate politically in destination cities.
0 if assigned to control, u is the error term, and X is a Regarding context, we relied on the community-level
vector of baseline covariates, included to improve statis- embeddedness of our partner organization and field
tical precision. All specifications control for gender, age, team (a) to be confident ex ante that the intervention
was contextually and culturally appropriate; (b) to
religion, caste, education, income, marital status, length
ensure that the registration intervention, as well as
of residence in the city, and homeownership status.27 common knowledge of it, would not invite pushback—
Where specified, we also included pretreatment meas- based on our partner’s relationships with local elites
ures of outcomes. T1 analyses use Huber–White robust and past experience running and publicizing such cam-
standard errors. paigns; and (c) to have in place a set of protocols
For the T2 analysis, we employ weighted least capable of providing immediate feedback on potential
squares regression: unanticipated events on the ground. In addition, we
sought the permission of the Election Commission of
Yij ¼ a þ b T2ij þ X0 i λ þ δs þ uij : (2) India as well as community leaders in each slum colony
prior to beginning work there.30 We further vetted the
Here, Y is the outcome of interest, T2 is a binary
indicator taking 1 if individual i in cluster j was assigned
28
See https://osf.io/7vtqh.
29
T1 control group participants remained free to submit applications
26
The communications specify that migrants have moved from other for voter ID cards at any time, and many did, speaking to the
parts of India and possess full constitutional rights to vote in destin- routineness of this process.
30
ation areas. The Election Commission has itself promoted migrant voter regis-
27
The income covariate was Winsorized to address significant out- tration in their place of “ordinary residence” (Tata Institute of Social
liers (see Online Appendix H). Sciences 2015).
1137You can also read
